Empty Words in Psychiatric Records: Where Has Clinical Narrative Gone?

Thomas G. Gutheil, MD;Robert I. Simon, MD;

Careful medical documentation is a primary factor in patient care, risk management, and liability prevention.^2-6 Such documentation represents a contemporary record of the flow of patient care; thus, it can be extremely helpful to the practicing clinician and can be an important factor in error prevention.

Unfortunately, although a number of forms of documentation may simulate valid recordkeeping, in actuality these are hollow shells of verbiage rather than appropriate chart content. Here we address some of these problems of meaningless phrasing, empty shells, and template-distorted recording in an attempt to improve clinical documentation for both clinical care and risk management.

It is important to face the grim realities that underlie the issue of documentation: documentation is hard, boring, time-consuming, and unloved by almost all practitioners. In the here and now hurly-burly of clinical activity, it is hard to appreciate the value and long-term benefits of a solid, useful chart. Thus, it is easy to understand the seductive lure of any shortcut (eg, abbreviations) that promises-no matter how speciously-to lighten the load. While some abbreviations may be well understood and quite useful (MMPI-III is much more convenient than the full title written out), the examples below reveal abbreviations that do not accomplish a valid clinical purpose.

The lure is especially seductive given the striking (perhaps appalling) strictures on the time that doctors are currently allocated to spend with patients; under managed care in the modern era, the 5-minute medication check is not uncommon. The temptation also to compress the note taking can become extremely strong.

Listen to the podcast on this topic

Psychiatric Records

Modern documentation occurs in a novel electronic context. The electronic health record (EHR) was promised by the Health Information Technology for Economic and Clinical Health (HITECH) Act of 2009 to “save 100,000 lives a year and $77 billion annually.” In reality, no savings have been produced and medical errors have not been reduced nor have lives been saved; indeed, errors have been added.^7,8 Moreover, since the EHR infrastructure is fundamentally insecure, 513 separate data breaches from 2005 to 2011 have affected more than 20 million patient records-and this number is limited to those actually reported.⁷

One of the most common documentation shortcuts that offers an illusion of appropriate mental health assessment is the famous triad: no SI/HI/CFS. This, of course, stands for “no suicidal ideation, no homicidal ideation, contract for safety.” Although these abbreviations suggest that at least some questions were asked, this handful of letters does not represent an adequate suicide or homicide risk assessment of the patient, especially in an evaluation in which these particular concerns arise. These 7 letters are so deeply ingrained in clinicians that in a recent chart review, we found the statement “continuing suicidal ideation” in the history section but noted, in complete contradiction, “no SI/HI” in the mental status segment.

Another shorthand expression is WNL, “within normal limits.” While this abbreviation might be adequate for narrow aspects of a physical examination or certain highly specific laboratory values, it is difficult to attribute any meaning to this acronym when applied to mental status issues that require at least some elaboration. What does it mean to say one’s fund of information is within normal limits? How could any clinician possibly identify what is “normal” about particular mental status entries without more specific content? If one could identify “normal,” what would it mean? Two variants on this latter topic include “NML,” meaning normal, and “unremarkable,” perhaps the vaguest of them all.

“Pt. left AMA” without further content fails to offer subsequent treatment providers a clue as to the reason: disagreement with the treatment plan, rage at treatment providers, inability to pay, lack of insight into illness? Each of these precipitants and others not listed would be quite important for appropriate responses by subsequent clinicians.

Yet another shortcut is NKDA, “no known drug allergies.” Although it may be reassuring to indicate that no allergies are known, this acronym does not answer the question of whether the issue was actually raised by the interviewing clinician or whether it is merely a statement of that clinician’s ignorance of a potentially important clinical factor.

Another regrettably generic phrase is “by history.” This phrase is often used to refer to past significant events of a clinically relevant nature, such as past suicide attempts, incarcerations, and involuntary hospitalizations. Because it lacks any reference to a particular source, it is often unclear whether “by history” refers to the history that the patient is providing by self-report or whether the datum is recorded in some medical documents; if the latter, what are those documents?

According to DSM, the abbreviation NOS, “not otherwise specified,” represents a presumably appropriate inclusive attempt to describe a patient whose complex and idiosyncratic presentation offers a diagnostic pattern that simply fails to fit into the convenient pigeonholes provided by the diagnostic manual. In some charts, the abbreviation suggests that the writer of the entry cannot be bothered to be specific or to commit to a particular diagnosis. We have heard a story of a forensic psychiatrist who when testifying as an expert witness in an insanity context, refers to the diagnosis of all his examinees as “psychosis NOS” in order to forestall being cross-examined on the listing of necessary diagnostic features from the manual.

A puzzling entry often seen in the nursing notes in assessments in psychiatric inpatient charts is the comment: “skin warm and dry.” Neither the psychiatric nor mental status significance of this entry is clear, but it may merely suggest that the nurse in question has just transferred from a medical-surgical setting and has internalized this rote phrase from frequent usage in that earlier context.

“Patient pleasant on approach” begins a number of nursing notes that we have seen. “Pleasant” is sufficiently vague to embrace everything from cooperativeness to frank hypomania; but does “approach” mean seeking out the patient for an evaluation or pass-ing within the patient’s social orbit in a random encounter? Or does it refer to the patient approaching staff?

“Non-contributory,” as in “family history non-contributory,” gives the illusion of performing some kind of “rule-out” function but is sufficiently generic not to be useful. Can one seriously maintain that a patient’s family history does not contribute in any way to the patient’s condition, symptoms, or psychological makeup? Here again, the value judgment inherent in that term should have been replaced by a narrative about what the history actually revealed.

Note that the expression “rule out,” which indicates a task to be done in the future, is often misinterpreted by attorneys as a statement about the past, namely “that diagnosis has been ruled out.” Some clinicians suggest using “consider” as a preferable alternative (R. Sadoff, MD, personal communication, October 5, 2011).

Finally, in addition to empty words, one can find a chart with no ending: even after years of treatment notes, the record simply stops-blank. Did anyone notice that the patient was no longer coming, or was this an agreed-on (although completely undocumented) termination? If the former, were any efforts made to reach the patient? Was the patient transferred? Did the patient die?

Psychiatric records depend on narratives to provide a living and comprehensive picture of the patient. This goal is thwarted by the increasing use of checklists and simplistic templates that leave record notes stripped of personal meaning and context. Some authors suggest that the very appearance of this template-driven recoding is influenced by the DSM itself, which promotes a checklist approach to diagnosis.⁹

Such devices are a mixed blessing: on the one hand, they may act as prompts to ensure that certain questions will be asked and certain topics considered; on the other hand, in the common form of a chart page teeming with many, many tiny checkboxes, such devices completely obscure any narrative picture of the patient. In addition, a box or two may be unnoticed, and thus a number of important areas of inquiry are overlooked. Furthermore, the template designer may not be a clinician, and therefore a clinically important area may simply have been omitted. If some element is omitted, the computer blindly responds and fills in the blanks, often redundantly, as seen in the following true example: PSYCHIATRIC HISTORY: History of depression, history of anxiety. History of anxiety, history of depression. No psychiatric history.

Another major fault of some computer-generated templates is their tendency toward redundancy, repeating some or all previous entries. For example, a note on the fifth hospital day is preceded by all the notes of the previous 4 days and the sixth day’s note has the previous 5. Keeping pace with the evolution of the clinical case becomes increasingly daunting; simultaneously, the redundancy fails to allow the clinician to obtain an evolutionary picture of the patient as a person undergoing changes.

Other forms of redundancy also appear: Social history: Denies alcohol abuse, denies tobacco abuse, denies drug abuse: patient smoked tobacco but quit at least 10 years ago. Patient consumes alcohol socially; lives at home with family. Denies alcohol abuse, denies tobacco abuse, denies drug abuse. Patient smoked tobacco but quit at least 10 years ago. Patient consumes alcohol socially; lives at home with family.

Documentation remains very important, but practitioners should avoid deluding themselves into thinking that checking off forms and boxes (although sometimes necessary) constitutes adequate and sufficient documentation. This caution should be heeded by all clinicians and taught to trainees in all the relevant disciplines.

The core risk management issue is the fact that stock entries usually suggest that the proper assessment or intervention was not done, especially when unaccompanied by narrative that supplies the necessary context. It is too easy for a fact-finder or subsequent decision maker-be it in a liability claim, a board of registration complaint, or an ethics complaint-to see the stock phrases as indices of a rote mentality that fails to take into account the individual clinical needs of the particular patient in question.

One of the most important risk management functions of documentation is capturing the use of clinical judgment: exercise of clinical judgment is an antithesis of negligence. But judgment cannot be captured through a host of checkboxes, especially when the checklists are designed primarily for recording individual symptoms or signs rather than complex mental processes.

If writing abbreviated notes is such a bad idea, is there really a need for documentation? Lawyers sometimes take the position that “bad documentation is worse than no documentation.” While one can understand the lawyers’ perspective and their concerns about what to bring to court, clinicians need to remind themselves that documentation is not primarily for lawyers: the main concern is always the direct effect on patient care. If the clinical picture of the whole patient is absent or blurred, care suffers.

At least 1 hospital has eschewed the use of an electronic medical record for psychiatric emergency admissions. In part, this move is justified by the fact that the existing data entry system allows only a few computer characters to record historical information. For psychiatric patients arriving at the emergency department, there was simply not enough space for narrative entries. A new version of the entry system allows narrative to be included for these patients.

Historically, psychiatric material has been presented in narrative form. Similarly, patients tell us their histories in the same format-their narratives provide the chronology of their experiences as persons. Moreover, the vocabulary, syntax, grammar, and associative connections in the patients’ own stories tell us far more about them as persons than a simple listing of the facts and events that mark their life experiences.

While today the definition of a formulation has taken various forms, an old version popular in Boston was the triad of premorbid personality, external precipitating stress, and type of reaction. Although this model leaned heavily on an event-based theory of the emergence of psychopathology, it contained a narrative vision of what the patient experienced and afforded a holistic view of the patient.

A narrative framework is still needed to capture and convey what a patient is all about. This model should be taught as part of clinical training in all disciplines. Since writing clinically relevant narratives is not an inborn skill, and since all narratives are not uniformly useful in providing a valid picture of a patient, composing clinically meaningful narratives ought to be a seminal component of all training programs.

References:

1. Shakespeare W. Macbeth. http://shakespeare.mit.edu/macbeth/full.html. Accessed November 14, 2011.
2. Gutheil TG. Paranoia and progress notes: a guide to forensically informed psychiatric recordkeeping. Hosp Community Psychiatry. 1980;31:479-482.
3. Gutheil TG. Documentation and sexual response. Synapse. Nov-Dec 1995:3.
4. Gutheil TG. Fundamentals of medical record documentation. Psychiatry (Edgmont). 2004;1:26-28.
5. Simon RI. Suicide risk assessment forms: form over substance? J Am Acad Psychiatry Law. 2009;37:290-293.
6. Simon RI, Shuman DW. Therapeutic risk management of clinical-legal dilemmas: should it be a core competency? J Am Acad Psychiatry Law. 2009;37:155-161.
7. Blackwell T. Electronic prescriptions do not reduce errors: study. National Post. March 29, 2011. http://www.nationalpost.com/Electronic+prescriptions+reduce+errors+study/4524330/story.html. Accessed November 14, 2011.
8. Spruiell GL. Keeping tabs on tabulators: patient identification transformation. The American Psychoanalyst. In press.
9. Gopal AA, Bursztajn HJ. DSM biases evident in clinical training and courtroom testimony. Psychiatr Ann. 2007;37:604-617. http://www.psychiatricannalsonline.com/showPdf.asp?rID=23800. Accessed November 15, 2011.